JPS5932037A - Halftone threshold generation apparatus and method - Google Patents

Abstract

(57) [Summary] This bulletin contains application data before electronic filing, so abstract data is not recorded.

Description

DETAILED DESCRIPTION OF THE INVENTION The present invention provides the ability to
A circuit for producing several types of halftone images, including low resolution, flicker-free halftone output for CRT displays, and high resolution halftone output for business-quality hardcopy. .

One common procedure in printing systems for editing images and text is to scan the data, edit the material using a CRT display, and print the edits. The input data can be character-encoded text read from magnetic media or a Rascou manual scanning device (rasternput 5canner = RIs).
) can be a continuous tone or halftone image scanned through the image. The RIS output may be, for example, 8 bits per pixel gray scale data.

There are different halftone patterns for CRT displays and hardcopy printing. Halftone patterns for CRTs have black and white pixels distributed relatively uniformly over an area and have a low resolution, eg, 80 pixels per inch. For ノ＼-F copying, pixels are classified into white and black pixels of various sizes, e.g. 48 pixels per inch.
The resolution is high, such as 0 pixels.

Generation of these different halftones requires two programs or, for high speed, two circuits.

It is an object of the present invention to provide a single circuit that can effectively generate both vacuums from the same scanned image data base.

Another problem with the CRT half-line pattern is that it introduces undesirable flicker into CRTs using interlaced rasters.

More specifically, in the system described here,
All CRT displays are generated at approximately 15 displays per second, and at that rate there is no noticeable flicker.
is confirmed. To correct this, odd rasters are used to generate one display and even rasters are used to generate the next display. As a result, the rate of display is 30 per second, and no flicker occurs.

The above configuration can be used for black and white text or line art, but for halftone images, if the bits chosen to represent the gray tones are broadly located on an odd (or even) raster. , a linker will still occur. In this case, flicker may occur in certain parts of the halftone image. The solution is to use a half-one pattern that ensures that the number of black pixels on odd and even lines is approximately equal for all gray levels.

The circuits and halftone patterns used to achieve these objectives will be described below with reference to the drawings.

FIG. 1 is a double spiral half-one pattern suitable for xerographic ink printing.

FIG. 2 shows the prior art dither ordered
r ordered) halftone pattern.

FIG. 3 shows a prior art two-level ordered dither (tw
o 1 level ordered di-t
h e r) It is a halftone pattern.

Figure 4 shows the night tour. tour) is a halftone pattern.

Figure 5 shows the night ordered dither (kntg-h, t
This is a bar 7 tone pattern.

Figure 6 shows the night/kn dither
light dither) halftone pattern.

FIG. 7 is a conceptual diagram of a circuit for generating these patterns.

FIG. 8 is a schematic diagram of a circuit for generating these buffers.

FIG. 9 is a group of color tone display curves showing changes in contrast and density.

FIG. 10 is a program that generates a threshold level for generating the curve group shown in FIG.

After the text and image data has been reviewed or loaded into the system, the next step in the editing process is to move the image and text to the first page on the display and adjust the image size, position, density, contrast, etc. and position all the text that appears on the page. When the page is in its final form, it can be printed, stored, transmitted, etc.

Displaying image data on a screen and printing it requires a halftone pattern.

The pattern used for printing in this system is a double spiral 8x8 matrix with gray levels varying from 0 to 62, as also shown in FIG. As is well known, as the image data of a document changes from white to black, the size of the white portion becomes smaller and the size of the black portion becomes larger.

Although this pattern produces good results for xerographic ink printers, a system that distributes pixels more evenly across the screen area is better for CRT displays.
more suitable.

A prior art dither pattern used in flat panels is shown in FIG. This type of pattern distributes pixels over an area, but does not have sufficient levels.

To provide 16 levels, four of these dither patterns can be combined into a two-level ordered dither pattern as shown in FIG.

However, this cannot be used in systems that use interlaced rasters on CRTs because, at a certain density level, it will cause flicker. For example, assuming a density level of 4, 1.2,3°4 of the pixels will be black and the rest will be white. All displays on even rasters have 4 black pixels (row O and 21 pixels 1.2, 3.4), while on all odd rasters (rows 1, 3) the pixels are all white.

As a result, a linker is confirmed.

Although the night tour halftone pattern shown in Fig. 4 has the characteristic of uniform distribution like the pattern in Fig. 3,
Produces minimal flicker on Hiektsu Lasku's CRT display. However, it has only 9 levels.

The night ordered dither pattern shown in FIG.
6 matrices.

In this pattern, corresponding pixels in each 3×3 quadrant are separated. Since there are three rows, the difference in the number of black pixels for one matrix of even and odd scan amounts is never greater than one.

For larger numbers of levels, the night/night dither pattern shown in FIG. 6 is used. This is nine 3x3
A 9x9 matrix of night tour segments, each of which is itself a night tour (rather than ordered dither as shown in Figure 5). In FIG. 6, the first 29 levels are filled in to show the pattern. The imbalance in this case is minimal (13
16) and are sufficiently distributed over the area to produce a flicker-free display.

As described above, the night ordered pattern has three effective characteristics. This is because the pattern distributes the black pixels evenly across the matrix, and the even and odd scan amounts provide sufficient pixel balance to produce the hardcopy pattern shown in Figure 1, as explained below. This means that it is generated by the same circuit that is used to do this.

Circuits for generating both halftone patterns are shown in the schematic diagram of FIG. 8 and the conceptual diagram of FIG. 7, respectively.

As shown in Figure 7, 16 x 16
Random access memory (RA) with a maximum capacity of words
M) is loaded with darkness values for either the whole of FIG. 5 or the matrix of FIG. Here, the fifth
Assuming the 6x6 matrix in the figure, then the first six rows of the first six columns of memory are loaded. In other words, the setting position is 0. 1. 2. 3. 4. 5.
16.17.18.19°20, 32.33, etc.

Also, both scan counters are cents modulo six. Finally, in hardware, a "forward scan counter" is incremented after each pixel, a "cross scan counter" is incremented at the end of each rask, and the output of each "screen threshold data" word is incremented after each pixel. , is compared against the corresponding "gray image data" in the comparator.

In this case, location 0 in FIG. 7 has a value as an element in row 0 and column 0 in FIG. 5, indicating a threshold level of 21. Then, as a series of pixels in Lascou are compared,
A series of halftoned binary bits is output from the comparator as "screened image data."

First step from 6×6 night ordered dither pattern generation
To change the operation to the 8x8 double spiral pattern shown, the first eight rows of the first eight columns are loaded with the appropriate threshold for the new pattern and the counter is changed to modulo eight.

By loading the appropriate threshold in RAM and resetting the counter to the appropriate value, this circuit can
It is clear from the above description that it can be used to generate any desired number of halftone patterns of up to pixel size. In fact, a 1×1 matrix can also be used. This is the threshold for black and white text and line art. In use, the darkness value is loaded in the first location and all pixels are compared to it.

The actual schematic diadem of this circuit is shown in FIG. When writing is enabled by the [-load screen counter O] line, the transverse scan counter 7
1 is loaded with a negative value through the [screen size OO~03] line. The counter 71 operates from that time until it counts 0, and when it reaches 0, it generates a [Screen Carry-0] signal (overflow signal). During that count, [screen address 00-0
The four line outputs of 3) are used to address the screen RAM 73. Counter 72 similarly operates to generate the other four address bits;
As a result, all 256 words of RAM are addressable.

During operation, the forward scan counter clocks each pixel (clock).
is incremented by and reloaded at the end of each matrix. Meanwhile, the transverse scan counter is incremented at the end of each line and reloaded at the end of each matrix.

[Dark values 00-07] can be loaded by [Screen write clock co-line, 8 outputs [Threshold 00-07]
07] are connected to comparators 75 and 76 together with gray image data [gray pixels 00 to 07]. The output line of the comparator is called the "screened pixel".

Gate 77 causes transverse scan counter 71 to increment at each end.

When a line is finished and the last matrix row is finished (this is when the counter 71 overflows and generates the [Screen Carry-0] signal), it is the 0th row of the matrix, and each At the start of a page, gate 78 restarts transverse scan counter 71. If counter 72 overflows or there is a screen size signal or line change, gate 79 restarts forward scan counter 72 at column 0]
− to make.

From the previous discussion, it can be seen that the threshold levels will be evenly distributed over the entire range of inputs. For example, FIG. 5 assumes that the threshold levels exhibit a uniform density distribution from 1 to 36. This is represented by the shape of curve number ■ in FIG. Threshold number 1 in FIG. 5 is assigned a grayscale value of 1, and threshold number 2 is assigned a grayscale value of 2, as shown by curve 1 in FIG.

The same applies below. In this way, the grayscale value is
Compared against a set of averagely distributed darkness values in a comparator, the output density and contrast of the copy will be equal to that of the original.

Additionally, the system allows an operator to adjust various threshold levels using interactive software. The need arises if the operator determines that the scanned image does not have sufficient contrast, density, etc. This software allows the operator to adjust these parameters before printing. Ideally, both the CRT and the printer would change identically so that the operator could
You can see the adjusted image.

For example, as shown in curve 3 in Figure 9, all threshold numbers below 8 are assigned a grayscale value of 0 to increase contrast, and all threshold numbers above 28 are assigned a grayscale value of 36.
is assigned a value of As a result, pure black and pure white areas will appear on the copy with a higher than normal probability, so that the contrast of the copy will be stronger than that of the original. Similarly, the average density can be changed by shifting curve 1 to the left or right. In fact, any shape of curve can be created with suitable software.

FIG. 10 is a program showing an example of the software, which generates a group of curves similar to FIG. 9 corresponding to the night ordered dither pattern of FIG. 5. This program is a Me
It is written in Sa and provided to the software to allow the operator to control the tone display curves on the interactive hesis.

The program of FIG. 10 allows the operator to change both the amount and shape of the bias in FIG. 9 using a single parameter: picptr, density. After determining some local variables, the program determines that the halftone cells (scre-enRecord) are two-dimensional arrays filled with appropriate fillers. Next, rowTable and co 1 umnTa
b le is determined. The values of each pair of matrix elements in Table are the coordinates of the pixel in the halftone cell, rowTable and column-"
The order of the coordinates in I'a b le is the order in which increasing values should be placed in cells to create the 6×6 night ordered dither pattern of FIG.

To calculate the screen value, the program uses picp
First, check for 0 values in tr and density. (picptr, dens -1ty, not shown in Figure 10, is an integer parameter ranging from 0 to 71.) This is a special case, where the image data is treated as a line drawing and halftoned. Rather than being enlightened, it shows that people will be uniformly enlightened. If pi
If cp-tr, den, s ij, y are 0, then (scr-eenRecord) is loaded as a single value of 20. Otherwise, -1) icp t
r, de-nsity includes some non-zero positive values, forming a night-ordered dither cell.

If the value is less than or equal to 36, then
The increment between successive darkness values is picptr, dCnsi −
Effective m a x G r divided by ty
a) 'Value. To position the threshold at its correct night-ordered dither position, the first FOR loop converts columnTable and row
- Use T a b 1 e to locate the consecutive values in the halftone cell. The process starts from the θ value and the maximum value is [maxGrayVa 1ue-1]
substantially ends the threshold generation so that it is limited to . Since all screen arrays are preset to all 1's prior to entry into the FOR loop, all pixels not addressed by the FOR loop remain set to their maximum value.

picptr, if the screen value occurs when density is greater than 36, picptr.

This is done in contrast to the method in which the value is determined when the density is 36 or less. The increment between successive darkness values is m
axGrayValue and picp-tr, densi
The effective maxGray divided by the difference from ty
Va 1. It is ue. The second FOR loop is cot-umn7” a b I e and rowT as before.
position the value to 5cleanRecord using ``able''. Also, by presetting all 5cre-enRecords to 0 before starting the operation to the FOR loop, all low-numbered pixels in the screen that are not addressed by the FOR loop become 0.

Although the present invention has been described based on specific embodiments, it will be apparent to those skilled in the art that various modifications can be made and constituent elements thereof can be replaced by equivalents without going beyond the scope of the inventive idea of the present invention. is obvious, and in addition, many changes can be made without departing from the spirit of the present invention.

[Brief explanation of the drawing]

Fig. 1 shows a double spiral halftone pattern suitable for xerographic ink printing, Fig. 2 shows a prior art dithered halftone pattern, and Fig. 3 shows a prior art dithered halftone pattern.
Level-ordered dither halftone pattern, Figure 4 shows night-time dither halftone pattern, Figure 5 shows night-ordered dither halftone pattern, Figure 6 shows night/night dither halftone pattern, and Figure 7 shows these patterns. 8 is a schematic diagram of a circuit for generating these patterns, FIG. 9 is a group of tone display curves showing changes in contrast and density, and FIG. 10 is a schematic diagram of a circuit for generating these patterns. This is a program that generates a threshold level for generating the curve group shown in FIG. 71: Transverse scan counter 72: Forward scan counter 73.74F screen RAM 75.76: Comparator? 7.78.79: Gate patent applicant Xerofukus Corporation agent Masu Kobori (and 2 others) FIG, 1 0] 2 3 FIG, J FIG, 5 H6, 6 Gumuis γ-L Tadashi H6, 9

Claims (1)

[Claims] 1. An m x m word RAM with n bits per word for temporarily storing dark values, and a memory for sequentially addressing locations in the RAM containing dark values along a certain matrix row. Two modulos for providing address information to said RAM, consisting of a first counter and a second counter that increments at the end of each line to change the row value from one matrix line to the next. A halftone darkness value generation device comprising a counter and a comparator for comparing each n-bit grayscale pixel to a respective threshold value to generate a binary output bit for each input that is a grayscale input. 2. Control means for enabling said apparatus to generate a plurality of halftone patterns, said control means comprising means for varying y and said R at a set of different darkness values.
A halftone darkness value generating device according to claim 1, comprising means for loading an AM. 3. Said control means is a computer, said computer further comprising program means for recalculating said darkness value by means of interactive operator control for varying the tone display curve. Halftone darkness value generation device described in section. 4, in the first y row of the first y column of mXm word, y row y
a first modulo y that addresses the RAM to increment through each row to load the threshold values of the half-column matrix and output the darkness value for each input that is a received grayscale pixel; Using a counter, at the end of each line, a second modulo y count to change the address operation from one line to the next, and a second modulo y count at the end of each line to generate a series of binary bit outputs. A method for generating a halftone pattern comprising the steps of: comparing scale pixels to an associated threshold level. 5. The halftone pattern of claim 4 further comprising the steps of varying the value of y and loading a set of different darkness values into the RAM to generate different halftone patterns. Generation method. 6. The method of claim 5, further comprising the step of interactively varying the tone display curve by recalculating and then loading another set of darkness values into the RAM. . 7. A method for assigning a darkness value to each element of a 3x3 element halftone matrix, comprising the following steps: Determine a set of nine darkness values that increase or decrease numerically, and write one row for each element of the matrix above, 6,9°4.3,
1. ．． Assign numbers such as 7, 8, 5.2, and assign the above dark value set to the above matrix in numerical order. 8. A method for assigning a darkness value to each element of a 6x6 element halftone matrix, comprising the following steps: Determine a set of 36 numerically increasing or decreasing darkness values, one row for each element of the above matrix, 21,33°13
, 23.35.15. 9. 1.25.11. 3.
27.29°17, 5.31.19. 7.24.3
6.16.22.34.14゜12, 4.28.10
．． 2.26.32.20. Assign numbers such as 8.30.18°6, and assign the above set of dark values to the above matrix in numerical order. 9. A method for assigning a darkness value to each element of a 3x3 element halftone matrix, comprising the following steps: Determine a set of nine numerically increasing or decreasing darkness values and assign these values to the above matrix according to the Night Tour Bakoon. 10. A method for assigning a darkness value to each element of a 6x6 element halftone matrix, comprising the following steps: A set of 36 numerically increasing or decreasing dark values is determined and these values are assigned to the matrix according to a night ordered dither pattern. 11. A method for assigning a darkness value to each element of a 9x9 element halftone matrix, comprising the following steps: Determine a set of 81 numerically increasing or decreasing darkness values and assign these values to the matrix according to a night tour pattern incorporating two levels.